A pot and resistive divider will work in exactly the same way. Any worry about the speed of the rising or falling edge is only going to show up in the harmonic content of the buzzer and as the fundamental frequency is quite high I would be surprised if you will here any difference.
What I mean is use the pot as an adjustable resistor (rheostat), use the wiper and one leg and let the other leg dangle. Sorry, I figured that was obvious given that he wanted to just "use a large resistor". I should have clarified.
I would imagine that to reduce the perceived volume of the piezo, you are going to have to increase the series resistance large enough to actually change the square wave to more of a reduced amplitude sawtooth waveform... so harmonics are not a factor. This is the least complicated way of attenuating the piezo output. The resistor divider way works as well, but is unbalanced.